Downward Sloping Swap Spread Curve

Question

After observing swap spreads in the market, I have noticed that the swap spread curve is downward sloping. Why is this? I have tried looking around the internet for answers, but have not found anything.

The only rational I could think of is that corporate issuers, who want to receive fixed on swaps to hedge their B/S interest rate exposure, have to enter in swap agreements with tenors of 5+ years (to match the maturity of their bonds), which drives the rate on the fixed leg of swaps down (and hence causes the swap spread curve to slop downwards).

Any input is appreciated.

EDIT: I am referring to US swap spreads.

Answer 1

If I look at the market I think this is mainly driven by the very nature of the long end investors of the swap curve. Compared to govi curves the swap curves provides a much better liquidity in longer tenors. Although we have seen a trend of bringing longer dated bonds to the market by government, too. Austria and Belgium are just two examples of these and Germany plans to add a new 50y benchmark bond. Now why is the swap curve downward sloping in the longer part causing the spread to be negative? On that part of the curve ALM investors are active and it is a huge market. These investors usually hedge their interest rate risk (in particular for DB plans, see also the new Dutch pension reform) using swaps. The bulk of liabilities has normally a duration of 25-35y. So effectively they are short a very long dated bond. To compensate they want to add duration to their portfolio to hedge this risk. Easiest way to do this is via receiving in the long end swap curve thus putting pressure on the long end swap curve.

Answer 2

This post is more related to EM markets, rather than developed markets (so could add some additional examples, to the already good DM examples given by @math above):

(i) In some countries (for example CZK prior to 2019), the Ministry of Finance preferred to issue shorter-dated bonds (up to 5 years), and there was less issuance of longer-dated bonds. As a result, pension funds and other type of funds, who wanted to go long duration, had no other choice but to receive longer-dated swaps. This caused the Swap curve to be "permanently" inverted from the 5y point up to the 10y point: it was just a "technical" flow problem.

(ii) The ECB has consistently failed to hit its inflation targets for the past decade. This has made the markets skeptical about inflation expectations in Europe (not just the Eurozone), and the markets tended to price in rate cuts for any sovereign European swap curves, where the central bank's policy diverged from the ECB policy of low rates (again, you could see this for example on the CZK curve, prior to the Covid19-driven change in CNB's policy).

(iii) During any risk-off scenario, fast money tends to receive emerging market swap rates between the 5y up to the 10y pillar (rather than receiving the short end): this can also make the curve inverted, whenever there is a flight-to-safety and at the same time the local central bank's rates are relatively high (which anchors the short-end of the curve)

Answer 3

Classically credit spreads widen with maturity, as the effect of default risk increases (and hence the extra yield required by investors) more rapidly for weaker credits.

In the “good old days” government bonds were AAA and OTC swaps were close to A-rated (reflecting the credit quality of the counterparts banks). Then came the Global Financial Crisis and investors were more sceptical of government bond ratings, while at the same time G10 swaps moved almost completely to clearing houses (which greatly diversified the counterparty risk).

In the US, the risk at the long end comes from a possible sizeable supply shock and extension of the maturity profile of Treasuries, as the deficit increases. Hence long-end Treasuries have been underperforming swaps.

In Europe, as others have commented, it is the use of long-dated swaps by pension funds which drives spreads via asset-liability matching. In times of market stress, stocks take a dive so the assets of funds decrease, while at the same time rates fall so the present value of liabilities increase. Funds need to hedge this by adding duration, and the most common way is to receive fixed on 20y-30y swaps (which of course only adds to the downward pressure on long rates). Thus you see 30y German bonds massively underperform 10y as the swap curve flattens sharply (while at the same time issuers become wary of long-end supply and government term premium picks up). We saw this post-Lehmans, during the sovereign debt crisis and most recently Covid: at the start of 2020 the swap spread curve on 10-30 was effectively flat, then dived in March and is now around 75% recovered.

Answer 4

If you look at models for swap spreads, the drivers for short, medium and long-end are quite different, i.e. there's a sort of implicit market segmentation hypothesis.

Re: Dimitri's comment on regs - you could look at the SDR volumes data to get an idea of whether long-end activity is declining.

EDIT: models I had in mind were for USD spreads.